How to argue in IELTS Part 3 and TOEFL Academic Discussion

These are argument tasks, not opinion tasks

IELTS Part 3 is a discussion — the examiner probes, challenges, and asks you to consider other views. TOEFL's Write for an Academic Discussion drops you into a thread where a professor and other students have already spoken.

In both, candidates answer as though the question were what do you think? And in both, the rubric is asking a harder question: can you build a case, and can you engage with the other side?

Jay Heinrichs makes the point that classical rhetoric was built on: argument is not the same as having views. It is a set of moves. Here are the ones that pay.

Move 0: control the tense — the one that decides everything else

In Thank You for Arguing, Jay Heinrichs recovers a piece of Aristotle that turns out to be the most practical thing in the whole of rhetoric. Every argument, Aristotle said, is really about one of three things, and each one lives in its own tense:

Past-tense argument (Aristotle's forensic) works out who did it, and threatens punishment. Present-tense argument (demonstrative) sorts the good from the bad, and ends with people bonding or separating. Future-tense argument (deliberative) asks what we should do — and, as Heinrichs puts it, promises a payoff. Aristotle liked the future best of the three, because it is the only one in which a decision can be reached.

Now look at what the examiner actually asked you

Should governments fund public transport? What should schools do about phones? How will cities change in the next fifty years? Nearly every IELTS Part 3 question, and nearly every TOEFL discussion prompt, is a choice question. They are deliberative. They live in the future.

And candidates answer them in the present tense — in values mode. Pollution is a serious problem. Governments are irresponsible. People should care more about the environment. That is a perfectly sincere set of evaluations, and it is an answer to a question nobody asked. You were asked what should be done, and you replied with what is bad.

The part that shows up in your score

Here is why this matters more than it looks, and it is the reason we put this move first. Your rhetorical mode decides which grammar you are even able to use.

You cannot deliberate about a future you are not certain of without reaching for conditionals and modals. The uncertainty is what produces the grammar. A candidate stuck in values mode is not merely answering the wrong question — they are rhetorically locked out of the structures their band depends on.

If you take one habit from this page: when a Part 3 question arrives, answer it forwards. What should happen, what would follow, what might go wrong with that. The examiner will hear an argument, and they will hear the grammar of one.

Move 1: concede and counter

The single most valuable move available in either task, and the most under-used.

You grant the strongest point on the other side — genuinely, not as a strawman — and then explain why your position survives it.

• Admittedly, free public transport is expensive, and it has to be paid for somehow. But the cost of congestion is also paid for — just invisibly, in time and in health.
• While it is true that technology isolates some people, the same technology is the only reason my grandmother speaks to her family every week.
• I take the point that stricter rules would be unpopular. The question is whether unpopularity is a good enough reason not to act.

Move 2: the concrete example

The fastest available fix for the most common cap in both tasks: the unsupported claim.

Technology has made people more isolated is an assertion. It is not an argument, and repeating it in different words does not make it one. Candidates stuck at 6 often have perfectly good ideas that are simply never supported.

A specific example is worth more than a general one, and a personal one is entirely acceptable in Part 3. It does not need to be a statistic — inventing statistics is worse than useless, and examiners are not fact-checking you. What matters is that the claim now has something under it.

There is also a language dividend. A concrete example forces past tenses, precise nouns and real vocabulary out of you, whereas an abstract assertion can be made entirely out of vague words. A vague answer cannot contain precise language, because there is nothing precise in it.

Move 3: define the term

Classical rhetoric calls this establishing stasis — agreeing on what is actually in dispute before arguing about it. It is the move that rescues a bad question.

Half of all weak Part 3 answers are the candidate arguing about a different thing from the one they were asked about, usually because the key word is vague.

• "Is technology making us less social?" — It depends what we mean by social. If it means seeing people face to face, then probably yes. If it means staying in contact, it has done the opposite.
• "Should governments fund the arts?" — It depends on which arts, and what funding means. A national museum and an individual painter are completely different arguments.

This does two things at once. It is genuinely good thinking, and it buys you time — you are speaking, fluently, on the topic, while you work out what you actually want to say. That is far better than a silence, and it is not padding: the distinction you draw becomes the structure of your answer.

For the TOEFL discussion specifically

The task gives you a professor's question and one or two student posts, and asks for a contribution in about a hundred words. Its particular trap is the one the rubric warns about: summarising instead of contributing.

1. Refer to what has been said — by name. Marco is right that… This is the single clearest evidence of contribution, and it costs you six words.
2. Add something new. Agreeing with an existing post and restating it is not a contribution. Either extend it, complicate it, or take a different position.
3. Concede and counter. Marco is right that remote learning saves time — although the students who benefit least are the ones who were already struggling to keep up. That single sentence engages, adds, and produces a concessive clause.
4. One concrete reason or example. In a hundred words, one is what fits.

Do not write an essay. There is no room for an introduction, and a hundred-word essay with a thesis statement and a conclusion has spent most of its words on scaffolding. Say the thing.

For the test, and for real arguments

For the test: concede and counter, support with something concrete, and define the term when the question is vague. Three moves, and between them they satisfy Task Response, Grammatical Range and Coherence at the same time.

For real life: these are not exam techniques. Conceding the strongest opposing point is how a person is persuaded of anything, and it is the move that separates an argument from a row. The specific example is what makes a case memorable in a meeting. And establishing what is actually in dispute before arguing is the single most useful thing anyone can do in a disagreement — most disagreements are about definitions and nobody notices.

Rhetoric has a bad reputation because it is associated with manipulation. What Heinrichs recovers is the older meaning: the craft of making a case that another person can actually accept. That craft happens to be exactly what the band descriptors are trying to measure — which is why practising it for the test is not wasted time.

The one thing in the book that does not transfer

Heinrichs is clear about what an argument is for. His test of success is not elegance and not even agreement:

In an exam, that goal is unavailable, and it is worth saying so plainly. The examiner is not your audience in Heinrichs's sense. They will not be persuaded, they will not act, and they genuinely do not care what you conclude — there is no criterion anywhere in the descriptors for having the right opinion. You cannot get your way, because the examiner wants nothing from you except language.